I would strongly recommend decent measurement system and a copy of Joe d'Appolito's "Testing Loudspeakers." This makes the design and optimization process much less random; doing a speaker project without the right tools is a sure path to spending piles of money and time, and most likely ending up with something that wasn't what you were hoping for. Joe's book will help you get good, valid, and useful data- understanding your tools is vital.
There's a lot of measurement options out there- for my current project, I'm using ARTA and VIRTINS with a sound card and a calibrated ICP mike, and CLIO Pocket. Very happy with both. I have an Omnic as well, and when it comes time to do a room setup, that will be my go-to. Any of those cost less than a RAAL ribbon.
Yes indeed. The education process is in motion (this can get pretty heavy!) I just read a piece by Joe d'Appolito on building a transmission line speaker. It is a very intriguing project. I'll look for more of his writing.
The measurement complexity is the reason I want to start with a kit and follow the general specifications provided. It also is the thinking behind the modular plans so that it can be modified in the future. Ultimately, the Swiss Army knife DEQX unit will allow me to adjust the crossover, phase, and EQ parameters and includes all the measurement tools needed to analyze it. I figure that building this initial project while continuing my course of study will position me well for when I make the investment in the more flexible DEQX.
Many articles worth reading: Linkwitz Lab - Loudspeaker Design
Also some very well respected speaker kits.
Also some very well respected speaker kits.
I found using battens with screws to hold the panels together was most forgiving.
Gluing panels means you have to be very accurate with your cuts. Even a couple of mm out means a gap in the wood as the glue wont fill the gap.
I get my panels cut by my local supplier on a vertical saw which gives very accurate cuts.
I leave sound deflecting panels just painted and cover the rest of the box in thin cord carpet.
Cabinet corners, grill, handles and a jack plug plate add to a professional look.
Gluing panels means you have to be very accurate with your cuts. Even a couple of mm out means a gap in the wood as the glue wont fill the gap.
I get my panels cut by my local supplier on a vertical saw which gives very accurate cuts.
I leave sound deflecting panels just painted and cover the rest of the box in thin cord carpet.
Cabinet corners, grill, handles and a jack plug plate add to a professional look.
Last edited:
The "Am I Crazy" question...
You must have noticed by now if 15 people respond you are getting 15 answers.
How big a room are you trying to fill, how loud do you like to listen, what type of music? Have you heard high efficiency speakers/ open baffle speakers/ systems with/without subs and what do you think about them? Do you only want 12 inch square speakers? When you see/hear something that makes you go "AW S**T" you've found where you want to go.
If you haven't noticed it yet at at the top of the Multi-Way speaker section there are a couple pages called stickies at the top. Check out the Ultimate Open Baffle Gallery sticky and the Systems Pictures & Description. That may give you some ideas.
You aren't crazy, but you are starting down a less traveled path that can be addictive. WARNING: It hasn't been a short walk for many on this site. The walk may last for years and years.
It's nice of you to join us not crazy people though .
Grant.
You must have noticed by now if 15 people respond you are getting 15 answers.
How big a room are you trying to fill, how loud do you like to listen, what type of music? Have you heard high efficiency speakers/ open baffle speakers/ systems with/without subs and what do you think about them? Do you only want 12 inch square speakers? When you see/hear something that makes you go "AW S**T" you've found where you want to go.
If you haven't noticed it yet at at the top of the Multi-Way speaker section there are a couple pages called stickies at the top. Check out the Ultimate Open Baffle Gallery sticky and the Systems Pictures & Description. That may give you some ideas.
You aren't crazy, but you are starting down a less traveled path that can be addictive. WARNING: It hasn't been a short walk for many on this site. The walk may last for years and years.
It's nice of you to join us not crazy people though .
Grant.
Indeed. I have really good speakers that will serve as the standard of comparison. I am embarking on the build as a great challenge.
Many opinions indeed, but gold is to be found within some! In fact, based on feedback, I've downgraded my initially over ambitious plan to a 2-way. I like this design and the kit saves a lot of guesswork:
https://www.madisoundspeakerstore.c...a-2-way-klang-ton-kit-pair-using-illuminator/
The room is about 12 x 17 feet, small but adequate and nicely isolated. I listen moderately to fairly loud, between perhaps 80 to 95db. Pretty diverse tastes, classic rock, all kinds of jazz, 70s/80s, singer/songwriters, some electronic, some symphony. Great female vocals is a real weak spot for me.
My last speakers were electrostats (ML CLX) so I have much experience with dipoles, although not open baffle conventional speakers. My current system is a sub sat with a very carefully integrated JL f113, so I have a great deal of experience with subs. I don't think full range is for me. I lean toward sealed boxes and I'm likely settling on that 2-way sealed box design but in a translam enclosure. I'm a pretty experienced woodworker with a decent shop, so a big part of this is doing a nice cabinet.
I expect that this is going to be a pretty lengthy commitment -particularly if this project comes out well. I think this could be a great next phase in my life-long audiophilia. I look forward to joining your community!
Last edited:
JRSystems,
A a few more projects (for which kits are available), and food for thought:
Ellam-FLEX
Which can be converted into a 3 way:
Ellam-FLEX-3W
Then another compact bookshelf:
Ellam98-mkII
And another:
https://sites.google.com/site/undefinition/carrera
Have fun!
A a few more projects (for which kits are available), and food for thought:
Ellam-FLEX
Which can be converted into a 3 way:
Ellam-FLEX-3W
Then another compact bookshelf:
Ellam98-mkII
And another:
https://sites.google.com/site/undefinition/carrera
Have fun!
JRSystems,
A a few more projects (for which kits are available), and food for thought:
Ellam-FLEX
Which can be converted into a 3 way:
Ellam-FLEX-3W
Then another compact bookshelf:
Ellam98-mkII
And another:
https://sites.google.com/site/undefinition/carrera
Have fun!
The Ellams are sweet! Great baffle and modular 3 way option. The ScanSpeak drivers seem like the way to go. I'm actually currently using the dbx 223 that he references. Interesting.
These look like they may be of interest.
No Audiophile's Reference system review / Selah Mejor Supreme
Grant.
No Audiophile's Reference system review / Selah Mejor Supreme
Grant.
Hey JRSystems,
If you have a DSP crossover, you can work your way though what is generally the most complex part of a conventional speaker system, the crossover. So that sets you is a great place to start.
I don't know how many speakers I have built. Too many - and according to my wife, those that remain in the house are still about 10 times as many as she would prefer
Analogue crossovers are great, but without a lot of math and experience they can be a bugger to get right. I love analogue,, but for various reasons I am a massive advocate of DSP crossovers -just look at my blog to get a handle on how silly I am about them.
Taking a very "engineering" approach - sorry that is what I am - here are a few things for you to consider.
- The bottom end can be "corrected" by DSP, but you will face limits of excursion and power handling that are a result of your bass driver(s) in the boxes they are in. Get your head around this through some modelling before you jump in. DSP correction is not a miracle cure.
- Make sure your midrange and treble drivers have a sensible "crossover region. Yes, you can correct these things in DSP, but if you are just getting kicked off, don't make things harder than you need. If as you state you are willing to pay for great drivers, then pay for great drivers that work well together and you will thank yourself later. Things to watch for are:
- Look out for "peakiness" in response (resonances) near the crossover region. You will hear this. And likely hate it. If a driver breaks up / has resonances in band, there is nothing that you can do to resolve this - bar crossing over lower / higher.
- Look out for choosing drivers that don't have a "common" crossover region, again, you can correct this with the crossover, but it will need power, and also force you into measuring things carefully.
- Time alignment. This is a whole topic...
If you don't get things time aligned, you will end up with a frequency response that just does not make sense" - until you work out that a time shift for a speaker driver is equal to a frequency dependent phase shift. This can create mayhem with your crossover.
DSP crossed systems need to consider this just as do conventional crossover systems.
How do I do this?
- You really do want a measurement mic and CRO to do it really well.
- Other people may have alternative approaches.
- There are a few great write ups - but fundamentally
- AT THE CROSSOVER FREQEUNCY IN QUESTION - set the mic at the measurement point in the far field.
- measure the tine delay / phase of the low freq driver
- Adjust the delay of the high frequency driver to match
- I use tone bursts so I can make sure I am not a whole cycle out... Not hard to do with a DSP crossover!
- Note *** - you also need to set the attenuation levels for each band to match one another.
If you get that right, then the "classic" crossovers simply work.
One other "engineering thing" that a DSP CANNOT fix is box shape.
- Again a whole science and art. Let alone emotive and perceptual thing.
- I have always liked the sound / soundstage of a minimally wide speaker, opting for greater depth than width.
- I am sure others will have a view. Fine. I am right and everybody else is of course wrong
- What do you like?
- Does this have an impact on the design?
- This is a big physical decision that will shape (!) your whole project!
back to the underlying question though:
- decent drivers in a decent box fed by a versatile DSP - if you can;t make it sound good, then you haven't finished.
- I will in fact be surprised if within half an hour of plugging it all together that you are not smiling a mile wide... just before you strart tweaking.
Have fun.
If you have a DSP crossover, you can work your way though what is generally the most complex part of a conventional speaker system, the crossover. So that sets you is a great place to start.
I don't know how many speakers I have built. Too many - and according to my wife, those that remain in the house are still about 10 times as many as she would prefer
Analogue crossovers are great, but without a lot of math and experience they can be a bugger to get right. I love analogue,, but for various reasons I am a massive advocate of DSP crossovers -just look at my blog to get a handle on how silly I am about them.
Taking a very "engineering" approach - sorry that is what I am - here are a few things for you to consider.
- The bottom end can be "corrected" by DSP, but you will face limits of excursion and power handling that are a result of your bass driver(s) in the boxes they are in. Get your head around this through some modelling before you jump in. DSP correction is not a miracle cure.
- Make sure your midrange and treble drivers have a sensible "crossover region. Yes, you can correct these things in DSP, but if you are just getting kicked off, don't make things harder than you need. If as you state you are willing to pay for great drivers, then pay for great drivers that work well together and you will thank yourself later. Things to watch for are:
- Look out for "peakiness" in response (resonances) near the crossover region. You will hear this. And likely hate it. If a driver breaks up / has resonances in band, there is nothing that you can do to resolve this - bar crossing over lower / higher.
- Look out for choosing drivers that don't have a "common" crossover region, again, you can correct this with the crossover, but it will need power, and also force you into measuring things carefully.
- Time alignment. This is a whole topic...
If you don't get things time aligned, you will end up with a frequency response that just does not make sense" - until you work out that a time shift for a speaker driver is equal to a frequency dependent phase shift. This can create mayhem with your crossover.
DSP crossed systems need to consider this just as do conventional crossover systems.
How do I do this?
- You really do want a measurement mic and CRO to do it really well.
- Other people may have alternative approaches.
- There are a few great write ups - but fundamentally
- AT THE CROSSOVER FREQEUNCY IN QUESTION - set the mic at the measurement point in the far field.
- measure the tine delay / phase of the low freq driver
- Adjust the delay of the high frequency driver to match
- I use tone bursts so I can make sure I am not a whole cycle out... Not hard to do with a DSP crossover!
- Note *** - you also need to set the attenuation levels for each band to match one another.
If you get that right, then the "classic" crossovers simply work.
One other "engineering thing" that a DSP CANNOT fix is box shape.
- Again a whole science and art. Let alone emotive and perceptual thing.
- I have always liked the sound / soundstage of a minimally wide speaker, opting for greater depth than width.
- I am sure others will have a view. Fine. I am right and everybody else is of course wrong
- What do you like?
- Does this have an impact on the design?
- This is a big physical decision that will shape (!) your whole project!
back to the underlying question though:
- decent drivers in a decent box fed by a versatile DSP - if you can;t make it sound good, then you haven't finished.
- I will in fact be surprised if within half an hour of plugging it all together that you are not smiling a mile wide... just before you strart tweaking.
Have fun.
Grant, thanks for the link. That is a terrific system and very similar to what I am ultimately shooting for.
Hey JRSystems,
If you have a DSP crossover, you can work your way though what is generally the most complex part of a conventional speaker system, the crossover. So that sets you is a great place to start.
I don't know how many speakers I have built. Too many - and according to my wife, those that remain in the house are still about 10 times as many as she would prefer
Analogue crossovers are great, but without a lot of math and experience they can be a bugger to get right. I love analogue,, but for various reasons I am a massive advocate of DSP crossovers -just look at my blog to get a handle on how silly I am about them.
Taking a very "engineering" approach - sorry that is what I am - here are a few things for you to consider.
- The bottom end can be "corrected" by DSP, but you will face limits of excursion and power handling that are a result of your bass driver(s) in the boxes they are in. Get your head around this through some modelling before you jump in. DSP correction is not a miracle cure.
- Make sure your midrange and treble drivers have a sensible "crossover region. Yes, you can correct these things in DSP, but if you are just getting kicked off, don't make things harder than you need. If as you state you are willing to pay for great drivers, then pay for great drivers that work well together and you will thank yourself later. Things to watch for are:
- Look out for "peakiness" in response (resonances) near the crossover region. You will hear this. And likely hate it. If a driver breaks up / has resonances in band, there is nothing that you can do to resolve this - bar crossing over lower / higher.
- Look out for choosing drivers that don't have a "common" crossover region, again, you can correct this with the crossover, but it will need power, and also force you into measuring things carefully.
- Time alignment. This is a whole topic...
If you don't get things time aligned, you will end up with a frequency response that just does not make sense" - until you work out that a time shift for a speaker driver is equal to a frequency dependent phase shift. This can create mayhem with your crossover.
DSP crossed systems need to consider this just as do conventional crossover systems.
How do I do this?
- You really do want a measurement mic and CRO to do it really well.
- Other people may have alternative approaches.
- There are a few great write ups - but fundamentally
- AT THE CROSSOVER FREQEUNCY IN QUESTION - set the mic at the measurement point in the far field.
- measure the tine delay / phase of the low freq driver
- Adjust the delay of the high frequency driver to match
- I use tone bursts so I can make sure I am not a whole cycle out... Not hard to do with a DSP crossover!
- Note *** - you also need to set the attenuation levels for each band to match one another.
If you get that right, then the "classic" crossovers simply work.
One other "engineering thing" that a DSP CANNOT fix is box shape.
- Again a whole science and art. Let alone emotive and perceptual thing.
- I have always liked the sound / soundstage of a minimally wide speaker, opting for greater depth than width.
- I am sure others will have a view. Fine. I am right and everybody else is of course wrong
- What do you like?
- Does this have an impact on the design?
- This is a big physical decision that will shape (!) your whole project!
back to the underlying question though:
- decent drivers in a decent box fed by a versatile DSP - if you can;t make it sound good, then you haven't finished.
- I will in fact be surprised if within half an hour of plugging it all together that you are not smiling a mile wide... just before you strart tweaking.
Have fun.
Terrific feedback! You validated much of my thinking. Thank you.
I do not have a DSP unit yet, but I plan on buying a very flexible DEQX unit before the end of the year. However, I first want to try my hand at a conventional two-way using a proven kit where the drivers, crossover, and box parameters are defined for me. That can then morph into a DSP electronic crossover with a bass cabinet or subs. Your points about driver parameters are definitely helpful -particularly the common crossover region. The DEQX has comprehensive measurement tools and I will likely spring for a high-end mic -so that is in the plans as well.
For the box, I am keen on doing a very rigid translam design of layered baltic birch. I think this design methodology is rigid, easily adapts to complex shapes, and looks cool as heck.
Your last line is the ultimate prediction. I foresee that first pair going live, a big smile on my face, and then the tweaking begins!
You don't even need a high-end mic, even some of the <$100 measurement mics will do fine (and measure more consistently than you'll get from even changing mic or speaker positions a few inches). I'd say you don't need DEQX, either, I don't see anything I'd want that the miniDSP 2x4HD and RePhase can't do, but who knows maybe DEQX has some hidden magic somehow.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.